Hi Andrea, this should be possible. However, if I remember correctly, there should be a check which ensures that a cell at the boundary does not get too distorted, i.e. if you use a curved boundary representation, then anisotropic refinement that would produce stretched elements along the boundary in combination with strong curvature of the boundary might be prevented. Is this the case for you?
Furthermore, typically a lot of anisotropic indicators are reset to isotropic if you use mesh smoothing. In that case, you should make sure that the allow_anisotropic_smoothing flag is set (in addition to anything you might want to use apart from that). I hope this helps, best regards, Tobias On Wednesday 30 March 2011 14:20, Andrea Mola wrote: > Hi everybody, > I am trying to anisotropically refine a 2d mesh, but I am finding some > problems with cells that happen to be on the boundary of the surface. > For instance, if the surface is composed of only one cell, it is not > possible to refine it anysotropically, as the > Triangulation<dim,spacedim>::execute_coarsening_and_refinement() method > executes only an isotropic refinement in such case. This happens even if > an anisotropic refinement flag has been set. > > >From what I have seen, I think that whenever asked to refine a cell with > > an edge on the boundary of the surface, in the direction that is > perpendicular to such boundary, the > Triangulation<dim,spacedim>::execute_coarsening_and_refinement() method > executes an isotropic refinement, regardless of the anisotropic flag > set. > Is there a specific reason why this occurs? And, is there a way I can > force an anysotropic refinement on an arbitrary direction on boundary > cells? > Thanks, > > Andrea > > _______________________________________________ > dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
